Description
Normal Computing is hiring an AI Inference Co-Design Engineer to develop algorithms and numerical methods that enable efficient transformer and diffusion inference on stochastic analog processing-in-memory hardware. The role co-designs algorithms with hardware and architecture teams, evaluates performance on real silicon or high-fidelity simulation, translates workload insights into hardware constraints, and iterates through simulation to silicon. Candidates should have strong knowledge of large-model inference, inference optimization, stochastic systems, and systems programming, with a track record of implementing ideas on real hardware.
